Mortality, Health Care Use, and Spending Patterns During South Korea's Trainee Physicians' Walkout

In February 2024, the South Korean government announced plans to increase medical school enrollment by 2000 students to address physician shortages, prompting opposition from the medical community. More than 90% of trainee physicians joined a nationwide walkout that continued through August 2025, but its systemwide implications remain unclear.

To examine the association of the trainee physicians' walkout with mortality, health care use, and spending patterns in South Korea.

This retrospective cohort study used nationwide health insurance claims data and a historical-control event-study design. Data from 200 teaching hospitals before and after the walkout in week 8 of 2024 were compared, using October 2023 to September 2024 as the exposed year and October 2022 to September 2023 as the historical control.

The primary outcomes were mortality, health care use, and spending per encounter measured weekly at the hospital level. The association between the trainee physicians' walkout and hospital-level outcomes was assessed using a group-time average treatment effect framework.

The analysis included 7 719 828 patients with inpatient admissions (3 520 242 [45.6%] aged ≥65 years; 3 883 073 [50.3%] female) and 41 891 392 with outpatient visits (17 552 493 [41.9%] aged ≥65 years; 21 992 981 [52.5%] female). The walkout was associated with a 0.008-percentage point increase in weekly mortality among hospital users (95% CI, 0.000 to 0.015 percentage points) and a 0.39-percentage point increase in 30-day mortality after hospitalization (95% CI, 0.24 to 0.54 percentage points), concentrated among admissions for general conditions. These associations were largest early in the walkout and attenuated over time. The walkout was associated with reductions in both inpatient and outpatient use, with a larger relative decline for inpatient care (decrease of 19.2 hospitalizations per 10 million people; 95% CI, -24.1 to -14.3 per 10 million people [-17.3%] vs decrease of 158 outpatient visits per 10 million people; 95% CI, -198 to -118 visits [-8.6%]), and with higher mean spending per hospitalization (US $301; 95% CI, $202-$399). Hospitalizations for general conditions showed the largest relative decline in service volume (-23.5%) and the greatest increase in mean spending per hospitalization (US $242; 95% CI, $108 to $375), whereas mortality, use, and spending for complex conditions remained stable. Little evidence of care shifting to nonteaching hospitals was found.

In this retrospective cohort study of teaching hospitals in South Korea, the trainee physicians' walkout was associated with small, short-term increases in mortality, reductions in hospital use, and higher spending per encounter, concentrated among admissions for general conditions. Outcomes for complex conditions were largely unchanged, suggesting that teaching hospitals prioritized high-acuity care. Rapid hospital and governmental responses may have helped limit harm and support resource allocation.